Unmanned aerial vehicles in wireless communications have caught great attention in unmanned aerial vehicle assisted communications and cellular-connected unmanned aerial vehicles. However, the unmanned aerial vehicle's advantage of strong line-of-sight channels also brings about severe and extensive interference between unmanned aerial vehicle and the terrestrial communication networks and between multiple unmanned aerial vehicles. This paper jointly optimises the unmanned aerial vehicle's three-dimensional trajectory and ground users' association schemes with the unmanned aerial vehicle and ground base station to reduce interference. In consideration of fairness, the authors formulate the optimisation problem as maximising the minimum average throughput among all ground users, a mixed-integer non-convex problem that is challenging to tackle. To ease the difficulties, the authors have the optimisation problem relaxed and propose a suboptimal solution in the framework of the block successive upper-bound minimisation algorithm. Furthermore, an improved block successive upper-bound minimisation based penalty algorithm is obtained to avoid the local maximum point problem. Numerical simulations have validated our schemes.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.